Abstract The physiological functions of many vital tissues and organs continue to mature after birth, but the genetic mechanisms governing this postnatal maturation remain an unsolved mystery. Human pancreatic β-cells produce and secrete insulin in response to physiological cues like glucose, and these hallmark functions improve in the years after birth. This coincides with expression of the transcription factors SIX2 and SIX3, whose functions in native human β-cells remain unknown. Here, we show that shRNA-mediated SIX2 or SIX3 suppression in human pancreatic adult islets impairs insulin secretion. However, transcriptome studies revealed that SIX2 and SIX3 regulate distinct targets. Loss of SIX2 markedly impaired expression of genes governing β-cell insulin processing and output, glucose sensing, and electrophysiology, while SIX3 loss led to inappropriate expression of genes normally expressed in fetal β-cells, adult a-cells and other non-β-cells. Chromatin accessibility studies identified genes directly regulated by SIX2. Moreover, β-cells from diabetic humans with impaired insulin secretion also had reduced SIX2 transcript levels. Revealing how SIX2 and SIX3 govern functional maturation and maintain developmental fate in native human β-cells should advance β-cell replacement and other therapeutic strategies for diabetes.
As of mid-April, Coronavirus Disease 2019 (COVID-19) has infected over 2 million people and resulted in nearly 150,000 deaths worldwide. This pandemic is principally a global health emergency, but the disease burden unavoidably impacts social dynamics and economic stability. While many countries have acted with swift and unified responses aimed at curbing exponential spread of the virus, pandemic interventions in the United States have been decentralized, particularly in the implementation of stay-at-home orders. To date, 8 states have not enacted statewide shelter-in-place measures and the country lacks a cohesive plan for lifting established stay-at-home orders. A thorough accounting of the evidence surrounding the impacts of stay-at-home measures can help guide both policy decisions and individuals’ actions. In this essay, we provide a multidisciplinary review of the effects and implementation of stay-at-home orders. We examine the epidemiological, health, economic, political, and legal issues relevant to assessing the costs and benefits of stay-at-home orders. We conclude that the evidence is in favor of implementing stay-at-home orders and maintaining them for the near future. The burden of these measures is less onerous than the health, economic, and political consequences associated with an acute spike in infections that would result from prematurely lifting aggressive public health interventions. To our knowledge, we present the most in-depth review of the evidence necessary to rigorously evaluate the full breadth of societal consequences associated with stay-at-home orders during the COVID-19 pandemic.
Abstract Aquagenic pruritus is a rare debilitating condition, which can be idiopathic, iatrogenic, or associated with systemic disease. In idiopathic cases, treatment can be challenging as options are limited and of variable efficacy. Here, we report the case of a teenage boy with refractory idiopathic aquagenic pruritus effectively managed with administration of β‐alanine supplementation, a treatment gaining traction in lay media but not yet reported in the medical literature. This report adds to the limited options published for treatment of idiopathic aquagenic pruritus in pediatric patients.
Photoreceptors are the most numerous and metabolically demanding cells in the retina. Their primary nutrient source is the choriocapillaris, and both the choriocapillaris and photoreceptors require trophic and functional support from retinal pigment epithelium (RPE) cells. Defects in RPE, photoreceptors, and the choriocapillaris are characteristic of age-related macular degeneration (AMD), a common vision-threatening disease. RPE dysfunction or death is a primary event in AMD, but the combination(s) of cellular stresses that affect the function and survival of RPE are incompletely understood. Here, using mouse models in which hypoxia can be genetically triggered in RPE, we show that hypoxia-induced metabolic stress alone leads to photoreceptor atrophy. Glucose and lipid metabolism are radically altered in hypoxic RPE cells; these changes impact nutrient availability for the sensory retina and promote progressive photoreceptor degeneration. Understanding the molecular pathways that control these responses may provide important clues about AMD pathogenesis and inform future therapies.
Article Figures and data Abstract eLife digest Introduction Results Discussion Materials and methods References Decision letter Author response Article and author information Metrics Abstract Photoreceptors are the most numerous and metabolically demanding cells in the retina. Their primary nutrient source is the choriocapillaris, and both the choriocapillaris and photoreceptors require trophic and functional support from retinal pigment epithelium (RPE) cells. Defects in RPE, photoreceptors, and the choriocapillaris are characteristic of age-related macular degeneration (AMD), a common vision-threatening disease. RPE dysfunction or death is a primary event in AMD, but the combination(s) of cellular stresses that affect the function and survival of RPE are incompletely understood. Here, using mouse models in which hypoxia can be genetically triggered in RPE, we show that hypoxia-induced metabolic stress alone leads to photoreceptor atrophy. Glucose and lipid metabolism are radically altered in hypoxic RPE cells; these changes impact nutrient availability for the sensory retina and promote progressive photoreceptor degeneration. Understanding the molecular pathways that control these responses may provide important clues about AMD pathogenesis and inform future therapies. https://doi.org/10.7554/eLife.14319.001 eLife digest Cells use a sugar called glucose as fuel to provide energy for many essential processes. The light-sensing cells in the eye, known as photoreceptors, need tremendous amounts of glucose, which they receive from the blood with the help of neighboring cells called retinal pigment epithelium (RPE) cells. Without a reliable supply of this sugar, the photoreceptors die and vision is lost. As we age, we are at greater risk of vision loss because RPE cells become less efficient at transporting glucose and our blood vessels shrink so that the photoreceptors may become starved of glucose. To prevent age-related vision loss, we need new strategies to keep blood vessels and RPE cells healthy. However, it was not clear exactly how RPE cells supply photoreceptors with glucose, and what happens when blood supplies are reduced. To address this question, Kurihara, Westenskow et al. used genetically modified mice to investigate how cells in the eye respond to starvation. The experiments show that when nutrients are scarce the RPE cells essentially panic, radically change their diet, and become greedy. That is to say that they double in size and begin burning fuel faster while also stockpiling extra sugar and fat for later use. In turn, the photoreceptors don’t get the energy they need and so they slowly stop working and die. Kurihara, Westenskow et al. also show that there is a rapid change in the way in which sugar and fat are processed in the eye during starvation. Learning how to prevent these changes in patients with age-related vision loss could protect their photoreceptors from starvation and death. The next step following on from this research is to design drugs to improve the supply of glucose and nutrients to the photoreceptors by repairing aging blood vessels and/or preventing RPE cells from stockpiling glucose for themselves. https://doi.org/10.7554/eLife.14319.002 Introduction Uninterrupted blood flow and an intricate and architecturally optimized network of photoreceptors, interneurons, glia, and epithelial cells are required for vision. The primary blood supply for photoreceptors is the choriocapillaris, an extraretinal fenestrated capillary bed. A layer of extracellular matrix proteins, Bruch’s membrane, lies adjacent to the choriocapillaris, and a monolayer of retinal pigment epithelium (RPE) cells divides Bruch’s membrane from the photoreceptors. The choriocapillaris, Bruch’s membrane, RPE, and photoreceptors function as one unit, with the choriocapillaris providing fuel for phototransduction, and Bruch’s membrane and RPE cells filtering and regulating the reciprocal exchange of oxygen, nutrients, biomolecules, and metabolic waste products between the circulation and retina. RPE also provide critical support for both photoreceptors and the choriocapillaris (Strauss, 2005) in large part by generating vascular endothelial growth factor (VEGF), a cytokine required for choriocapillaris development and maintenance (Kurihara et al., 2012; Le et al., 2010; Marneros et al., 2005; Saint-Geniez et al., 2009). Defects in this unit, including reduced choriocapillaris density, the presence of sub-RPE lipid-rich deposits, and RPE/photoreceptor dysfunction, are characteristic of age-related macular degeneration (AMD), a common vision-threatening disease whose prevalence is steadily increasing globally (Friedman et al., 2004; Wong et al., 2014). Several genetic and lifestyle risk factors have been identified but no cure exists (Bird, 2010). While current evidence suggests that AMD is a spectrum of closely related multifactorial polygenic diseases (Bird et al., 2014), 10 year clinic-based data from the Age-Related Macular Degeneration Study (AREDS; n = 4757) showed that the major risk factors include aging, severity of drusen (sub-RPE deposits), and RPE abnormalities (Chew et al., 2014). Early AMD is characterized by pigmentary changes and appearance of drusen (Gass, 1973; Pauleikhoff et al., 1990; Sarks, 1976; Wang et al., 2010). In most cases early AMD proceeds towards geographic atrophy (or 'dry' AMD), a condition defined by focal photoreceptor, RPE, and choriocapillaris loss and thickening of Bruch’s membrane with immunomodulatory proteins and lipids (Bird, 2010; Jager et al., 2008; Zarbin, 1998). While the primary defect could occur in Bruch’s membrane (Pauleikhoff et al., 1990; Bressler et al., 1990; Mullins et al., 2011), the choriocapillaris (Ramrattan et al., 1994; Spraul et al., 1996; Spraul et al., 1999), or photoreceptors (Sarks, 1976; Hogan, 1972; Sarks et al., 1988), most evidence suggests that it probably occurs in RPE cells. Granules enriched with lipid-rich residues, lipofuscin, accumulate normally in aging RPE cells (Bazan et al., 1990), but abnormal accretions are observed in patients with geographic atrophy in a band directly surrounding the lesion (Feeney-Burns et al., 1984; Holz et al., 1999; von Ruckmann et al., 1997). Lipofuscin renders the RPE more sensitive to blue light-induced damage (Rozanowska et al., 1995; Schutt et al., 2000; Sparrow et al., 2000), impairs RPE functions (Holz et al., 1999; Finnemann et al., 2002; Lakkaraju et al., 2007; Sparrow et al., 1999), and is potentially toxic for RPE cells (Schutt et al., 2000). The downstream effects of RPE loss are catastrophic, and result in choriocapillaris attenuation and photoreceptor degeneration in late stage AMD patients (Bhutto and Lutty, 2012; Coscas et al., 2014; Jonas et al., 2014; Lee et al., 2013; Sohrab et al., 2012; McLeod et al., 2009). Assuming, therefore, that the critical event of AMD pathogenesis occurs in RPE cells, how can the onset of the other early clinical manifestations of the disease in neighboring cells and structures be explained? There is a growing body of evidence that choroidal and retinal blood flow is reduced in AMD (Boltz et al., 2010; Remsch et al., 2000). We hypothesize that hypoxia, a natural consequence of aging microenvironments (that is exacerbated by obesity and smoking) (Blasiak et al., 2014; Chiu and Taylor, 2011; Morgado et al., 1994; Sagone et al., 1973), in RPE cells may be a central AMD risk factor based on the following lines of evidence: (a) RPE provide critical vasculotrophic support required for photoreceptor function (Kurihara et al., 2012); (b) Hypoxia alters lipid handling in other cell-types (Glunde et al., 2008; Santos et al., 2012; Semenza, 2009); (c) at least 40% of lipids in drusen are secreted by RPE (Cao et al., 2013); (d) the RPE secretome is sensitive to stress (Wang et al., 2010); and (e) in drusen rich zones of AMD patient eyes, vascular density is significantly reduced (Mullins et al., 2011). Therefore, hypoxia-mediated changes to the RPE lipidome and secretome could enhance lipofuscin accumulation and induce Bruch’s Membrane lipidization and thickening, thereby exacerbating RPE dysfunction, choriocapillaris drop-out, and photoreceptor dysfunction. This vicious cycle of events could accelerate progression of AMD. Hypoxia-inducible factor alpha subunits (HIF-αs) are the key transcription factors that mediate responses to hypoxia. Under normal conditions HIF-αs are constitutively expressed and targeted by von Hippel-Lindau protein (VHL) for ubiquitination and proteasomal degradation. VHL is inactivated at low oxygen tensions; this allows HIF-αs to translocate to the nucleus and activate a host of angiogenesis, glucose metabolism, erythropoiesis, and inflammation genes (Semenza, 2011). In this study we directly or indirectly hyperactivated HIF-αs in RPE by genetically perturbing components of the VHL/HIF/VEGF pathway using inducible and conditional gene ablation techniques. These manipulations altered lipid handling and glucose consumption of RPE cells, induced gross morphometric changes in RPE, reduced nutrient availability for the sensory retina, and promoted progressive photoreceptor atrophy. Understanding the effects of hypoxia on RPE metabolism, and learning how to control these effects, may provide insights for developing novel therapeutic strategies to treat retinal degenerative diseases. Results Choriocapillaris attenuation induces hypoxia in RPE and promotes photoreceptor degeneration Based on the hypothesis that choriocapillaris attenuation and prolonged hypoxia in RPE cells induces photoreceptor death/dysfunction, we set out to catalog the temporal and spatial manifestations of hypoxia in retinas of mice with severe choriocapillaris deficits (VMD2-Cre;Vegfafl/fl) and correlate these manifestations with any corresponding anatomical and functional changes in photoreceptors. Transgenic mice harboring human vitelliform macular dystrophy-2 promoter-directed cre (VMD2-Cre) (Le et al., 2008) were used to ablate Vegfa; severe choriocapillaris vasoconstriction is observed in adult Vegfa-cKO mice three days post induction (dpi) (Kurihara et al., 2012). The first signs of RPE hypoxia, including nuclear HIF immunoreactivity (Figure 1A), accumulation of the hypoxic probe pimonidazole in RPE (Figure 1A&B; white arrows), and activation of a known panel of hypoxia-inducible target genes, were observed six months post induction (mpi) in the Vegfa mutants (Figure 1—figure supplement 1&2). However, we cannot exclude the possibility that low-grade hypoxia may occur in RPE or other retinal cells earlier at subthreshold levels of detection. Hypoxia in RPE induced several defects including severely distended basal infoldings, accumulation of lipid droplets within RPE cells (Figure 1C; yellow arrows), RPE cell hypertrophy (Figure 1D), and dramatic and progressive thickening of Bruch’s membrane beginning at nine months post induction (Figure 1C red arrows; Figure 1E pseudo-colored blue). At 11 months post induction we also detected pigmentary abnormalities in fundus images (Figure 2A) and thinning of the photoreceptor cell layer (Figure 2B; red line) characteristic of photoreceptor degeneration. While RPE defects took months to manifest, defects in cone-driven pathways occurred within seven days of Vegfa ablation (Kurihara et al., 2012) and do not recover by 11 months post induction (Figure 2C and D, photopic). Surprisingly, rod-driven pathway defects were not observed until 11 months post ablation (Figure 2C and D, scotopic), suggesting that, for reasons that are unclear, rod photoreceptors are less sensitive to oxygen and nutrient deprivation than cones are. Figure 1 with 2 supplements see all Download asset Open asset HIF-α accumulation precedes the induction of AMD-like features in Vegfa-cKO mice. (A) HIF-1α, HIF-2α, and pimonidazole (Pim; green) are detected 6 months post induction in flat-mounted RPE/choroids from Vegfa mutants but not in littermate controls. ZO-1 (red) labels the cell boundaries. (B) The hypoxia probe Pimonidazole (Pim; green) is detected specifically in the RPE (white arrows) of cross-sectioned Vegfa mutant retinas (probe labeling is shown alone (left) and overlaid over brightfield images (right) to emphasize the retinal anatomy). (C) Electron micrographs of littermate control (left) and Vegfa-cKO RPE 11 months post induction (right). Dashed squares in left panels are magnified in right panels. Note the absence of choriocapillaris in Vegfa mutants, accumulation of lipid droplets (yellow arrows) in the cytoplasm, and thickening of Bruch’s membrane (red arrows). (D) Measured thickness values of the RPE of Vegfa mutant mice 11 months post induction (n=4) (see associated Figure 1—source data 1). (E) Electron micrographs of RPE/Bruch’s membrane from control and Vegfa-cKO RPE taken 9, 11, and 18 months post induction. Note the progressive accumulation of material in Bruch’s membrane (pseudo-colored light blue). Abbreviations: Pim=pimonidazole, INL=inner nuclear layer, ONL=outer nuclear layer (photoreceptor cell bodies), OS=photoreceptor outer segments, BI=basal infoldings of RPE cells, BM=Bruch’s membrane, CC=choriocapillaris, mpi=months post [RPE-specific] induction. Scale bars=20 µm (A), 50 µm (B), 1 µm C&E. https://doi.org/10.7554/eLife.14319.003 Figure 1—source data 1 Source data for Figure 1D. https://doi.org/10.7554/eLife.14319.004 Download elife-14319-fig1-data1-v3.xlsx Figure 2 Download asset Open asset Photoreceptor atrophy and dysfunction is observed in late stage Vegfa-cKOs. (A) Color fundus images of littermate control and Vegfa-cKO mice 7 or 11 months post induction (mpi). No obvious changes are seen after 7 months, but significant changes indicative of retinal degeneration are seen 11 months post induction. (B) Photoreceptor atrophy, determined by observing thinning of the outer nuclear layer (ONL; red vertical line) is seen in DAPI labeled cryosectioned Vegfa-cKO retinas 18 months post induction 600 μm from the optic nerve head compared with controls. (C) Full-field ERGs performed on controls and Vegfa-cKO mice 11 months post induction (n=6) reveal rod dysfunction in both the a- and b-waves (scotopic), and cone dysfunction in the b-wave and flicker response (photopic flash & flicker) in Vegfa-cKO mice. (D) Quantification of ERGs (see associated Figure 2—source data 1). *p<0.05, **p<0.01. Error bars indicate mean plus s.d. Scale bar=20 µm. https://doi.org/10.7554/eLife.14319.007 Figure 2—source data 1 Source data for Figure 2D. https://doi.org/10.7554/eLife.14319.008 Download elife-14319-fig2-data1-v3.xlsx RPE can induce choriocapillaris vasodilation Theoretically, RPE should be able to respond to hypoxia and improve circulation in the choriocapillaris by increasing basal VEGF secretion. To induce hypoxia we maintained 3D cultures of primary human RPE in 3% oxygen in a controlled chamber and analyzed the molecular and metabolic changes. The HIF target gene Vegfa was upregulated after exposure to low oxygen (Figure 3A) or upon addition of dimethyloxalylglycine (DMOG), an inhibitor of prolyl hydroxylase that leads to HIF stabilization (Figure 3B). DMOG also significantly stimulated basal VEGF secretion in a dose dependent manner (Figure 3C). To determine the physiological relevance of hypoxia-enhanced RPE-derived VEGF synthesis, we used VMD2-Cre to delete Vhl in RPE in vivo. Signs of [pseudo] hypoxia, i.e. HIF-α immunoreactivity and activation of known hypoxia-induced genes (including Vegfa), were observed three days post induction in Vhl-cKO mice (Figure 3D and Figure 3—figure supplement 1). The increased VEGF production from Vhl-cKO RPE correlates with significant and progressive choriocapillaris vasodilation based on ultrastructural examinations (Figure 3E and F). However, no vasodilation was observed in double (Vhl/Hif1a-dKO or Vhl/Hif2a-dKO) or triple (Vhl/Hif1a/Hif2a-tKO) knockout mice (Figure 3G) indicating that both RPE-derived HIF-1α and HIF-2α are mutually responsible for HIF/VEGF induced vasodilation. Collectively, these findings demonstrate that RPE can sense changes in oxygen/nutrient availability and respond by appropriately altering the vascular tone of the choriocapillaris. Figure 3 with 1 supplement see all Download asset Open asset HIF-αs induce dilation of the choriocapillaris in Vhl-cKOs. (A) Vegfa is upregulated in hRPE exposed to 3% oxygen. Data are the mean plus s.e.m. (n=5–6). (B–C) Vegfa mRNA (B) and VEGF165 protein (C—predominately from the basal surface when grown on transwells) is upregulated in a dose-dependent manner by DMOG for 24 hr compared with DMSO controls. Data are the mean plus s.e.m. (n=4–6). (D) Immunohistochemistry analyses reveal that VHL (green) expression is lost 3 days post induction in Vhl-cKOs, and HIF-1α and HIF-2α are upregulated in the nucleus of Vhl-cKOs RPE, but not in controls. (E) Measurements of the choriocapillaris from electron micrographs 0–28 days post induction (F) in untreated and Vhl mutants revealed progressive choriocapillaris vasodilation (n=4). (G) Choriocapillaris thickness values of Vhl-cKO, Vhl/Hif1a-dKO, Vhl/Hif2a-dKO, Vhl/Hif1a/Hif2a-tKO mice measured 28 days post induction (n=4). (See also associated Figure 3—source data 1 for panels A-C, E, and G.) Scale bars=20 µm. Error bars represent mean plus s.d. https://doi.org/10.7554/eLife.14319.009 Figure 3—source data 1 Source data for Figure 3A–C,E, and G. https://doi.org/10.7554/eLife.14319.010 Download elife-14319-fig3-data1-v3.xlsx Hypoxia induces structural changes in RPE and retinal degeneration Hypoxia-triggered choriocapillaris vasodilation may come at a significant cost for the retina. The accumulation of lipid droplets and material resembling glycogen is detectable within only three days post induction in the RPE of Vhl mutants (Figure 4A). Severely distended basal infoldings, thickening of Bruch’s membrane, and numerous lipid droplets (sometimes contiguous with subretinal extracellular spaces) are observed 14 days post Vhl deletion (Figure 4B; red). Measurements across the RPE from electron micrographs reveal significant hypertrophy (Figure 4C). While the double deletion of Vhl/Hif1a did not prevent the RPE defects in the Vhl mutants, the RPE cells of Vhl/Hif2a or Vhl/Hif1a/Hif2a mutants appeared unremarkable and were not hypertrophic (Figure 4D and E), suggesting that HIF-2α, as it is in other cell-types (Qiu et al., 2015; Zhao et al., 2015), is the pathological HIF isoform in hypoxic RPE. Figure 4 Download asset Open asset Dramatic and rapid-ensuing RPE defects observed in Vhl-cKO mice are dependent on Hif2a. (A) Electron micrographs of RPE cells from littermate control and Vhl-cKO mice 3 days post Vhl deletion. Regions marked with perforated white rectangles are in the lower panels. Note the intracellular accumulations of lipid droplets (a; red) and glycogen (b). (B) Electron micrographs of RPE cells from littermate control and Vhl-cKO mice 14 days post Vhl deletion. Intracellular lipid droplets (a), extrusion of lipid droplets into the subretinal space (b), and lipids collecting along the basal laminar surface of Bruch’s membrane and between RPE basal infoldings (c) are observed. (C) Thickness measurements from electron micrographs and reveal that RPE hypertrophy occurs from 0–3 day post induction timepoints, and then plateaus from 3–28 timepoints in Vhl mutant mice (n=5). (D) Electron micrographs of RPE from Vhl/Hif1a (left panels), Vhl/Hif2a (upper middle panel), and Vhl/Hif1a/Hif2a (bottom middle panel) mutant mice 14 days post induction. Note that lipid droplets (dark gray spheres, upper left panel) and material resembling glycogen (small punctate spots) are observed in Vhl/Hif1a-dKO, but not in Vhl/Hif2a-dKO or Vhl/Hif1a/Hif2a-tKO RPE) 14 days post induction. These data suggest Hif2a is responsible for the phenotype in Vhl mice. (E) Choriocapillaris thickness values of Vhl-cKO, Vhl/Hif1a, Vhl/Hif2a, Vhl/Hif1a/Hif2a mice measured 28 days post induction (n=4). (See also associated Figure 4—source data 1 for panels C&E.) Scale bars=5 µm (A), 1 µm (A’a & A’b), 2 µm (B), 0.5 µm (B’a, B’b, B’c), 5 µm (D). Error bars represent mean plus s.d. https://doi.org/10.7554/eLife.14319.012 Figure 4—source data 1 Source data for Figure 4C&E. https://doi.org/10.7554/eLife.14319.013 Download elife-14319-fig4-data1-v3.xlsx Devastating secondary effects resulting from RPE hypoxia were observed in the sensory retina. Photoreceptor degeneration (Figure 5A–C) and significant functional impairments of both rod and cone driven-pathways (Figure 5D—figure supplement 1A) were observed in VMD-Cre;Vhl and VMD-Cre;Vhl/Hif1a mice 28 days post induction but not in in Vhl/Hif2a (or Vhl/Hif1a/Hif2a mice; Figure 5C,E–G), (or in any of the relevant controls; Figure 5—figure supplement 1B and C). In advanced stages of the phenotype (>50 dpi), dramatic changes in RPE and the vasculature are observed consistent with retinal remodeling (Figure 5—figure supplement 2) (Marc et al., 2003). These findings imply that HIF-2α-mediated metabolic stress in RPE, which cannot be rescued even with a significantly dilated choriocapillaris, is enough to promote photoreceptor degeneration. Figure 5 with 2 supplements see all Download asset Open asset Progressive and rapid photoreceptor degeneration observed in Vhl-cKO mice is dependent on Hif2a. (A) Thickness measurements from electron micrographs reveal progressive thinning of the outer nuclear layer (ONL or photoreceptor cell bodies) from 0–28 days post induction timepoints in Vhl mutant mice (n=5). (B) Cryosectioned DAPI stained retinas from Vhl-cKO mice prior to induction (0 dpi; left panel), 7 dpi, and 28 dpi. (C) Quantified thickness values measured 600 μm from the optic nerve head of the outer nuclear layer in Vhl-cKO, Vhl/Hif1a, Vhl/Hif2a, and Vhl/Hif1a/Hif2a 28 days post induction (n=4) (See associated Figure 5—source data 1 for panels A&C.). (D) Full-field ERGs performed on Vhl-cKO and control mice 28 days post induction. (E) ERGs from Vhl/Hif1a, (F) Vhl/Hif2a, and (G) Vhl/Hif1a/Hif2a mutant mice 28 days post induction. ERG analyses reveal that normal photoreceptor function is observed in Vhl/Hif2a (F) or Vhl/Hif1a/Hif2a (G) mutant mice. *p<0.05, **p<0.01. For all ERGs n=6–8. Error bars indicate mean plus s.d. https://doi.org/10.7554/eLife.14319.014 Figure 5—source data 1 Source data for Figure 5A&C and Figure 5—figure supplement 1A–C. https://doi.org/10.7554/eLife.14319.015 Download elife-14319-fig5-data1-v3.xlsx Lipid handling is impaired in hypoxic RPE We next set out to identify the molecular mechanisms driving the hypoxic-mediated metabolic stress in RPE cells. Based on histopathological evidence of impaired lipid handling in hypoxic RPE we performed gene profiling for fatty acid metabolism genes from RPE/choroid complexes of Vhl-cKO mice. Acyl-CoA synthetase and Acyl-CoA dehydrogenase family genes were downregulated (Figure 6A and Figure 6—figure supplement 1A and B) but not in Vhl/Hif2a or Vhl/Hif1a/Hif2a mutant RPE (Figure 6B and Figure 6—figure supplement 1B). We also performed untargeted high-resolution mass spectrometry-based metabolomic analyses and observed abnormal levels of several long-chain saturated, unsaturated, and oxidized acylcarnitines in the Vhl-cKO mice (Figure 6C and Figure 6—figure supplement 2). Vhl/Hif1a mice exhibit dysregulated metabolomic profiles similar to Vhl mutants (Figure 6—figure supplement 2A and B), but normal levels of acylcarnitines and other metabolites were observed in Vhl/Hif2a (Figure 6—figure supplement 2C) and Vhl/Hif1a/Hif2a mice (Figure 6—figure supplement 2D). Collectively, these data strongly suggest that HIF-2α regulates lipid handling in RPE in vivo. Figure 6 with 3 supplements see all Download asset Open asset Defects in lipid metabolism in Vhl mutant RPE. (A) Summary of gene-profiling experiments for lipid metabolism genes in RPE/choroid complexes from Vhl mutant mice 3 days post induction (n=4). (B) Downregulation of lipid metabolism genes was also seen in Vhl/Hif1a mutants, but nominally in Vhl/Hif2a, and no gross changes were seen in Vhl/Hif1a/Hif2a mutants 3 days post induction (n=4). (C) Untargeted high-resolution mass spectrometry-based metabolomic analyses revealed that a group of acylcarnitines (AC) was progressively dysregulated from 3 to 14 days post induction (n=4–6) (see also associated Figure 6—source data 1). Box and whiskers plots are shown. Error bars represent maximum and minimum values. (D) Pre-treating hRPE with DMOG reduced the basal oxygen consumption rates (initial OCR – OCR after injection of RAA) when the cells were assayed in substrate limited media (2.5 mM glucose) and provided BSA control or palmitate conjugated to BSA (n=4) (see also associated Figure 6—figure supplement 3B). Error bars are the maximum and minimum values in panel C and mean plus s.d. in panel D. https://doi.org/10.7554/eLife.14319.018 Figure 6—source data 1 Source data for Figure 6C, Figure 6—figure supplement 2A–D, Figure 6—figure supplement 3B. https://doi.org/10.7554/eLife.14319.019 Download elife-14319-fig6-data1-v3.xlsx To test if HIF activation leads to altered lipid oxidation we monitored oxygen consumption in human RPE treated with a hypoxia mimetic, DMOG. Seahorse flux analysis revealed that human RPE cells in substrate-limited media oxidize exogenous lipids (palmitate) as an energy substrate (BSA control vs. palmitate control, Figure 6D). The ability of palmitate to increase oxygen consumption rates (OCRs) was validated by the addition of an inhibitor of the carnitine transport, etomoxir, which reduced oxygen consumption to BSA control levels (Figure 6—figure supplement 3A). Interestingly, treating the cells with low-dose DMOG for 48 hr prevented lipid oxidation (palmitate control vs. palmitate DMOG, Figure 6D and Figure 6—figure supplement 3B). These data suggest that hypoxic RPE alter their lipid handling behaviors and begin storing lipids in droplets, rather than utilizing them as an energy substrate. Glucose metabolism is disrupted in hypoxic RPE The presence of visible glycogen stores in Vhl-cKO RPE indicates that glucose metabolism may also be dysregulated. Using PCR arrays for glucose metabolism we determined that glycogen degradation genes were downregulated in Vhl mutant mice (summarized in Figure 7A). Furthermore, we observed that glycolysis-related genes were largely upregulated and TCA cycle genes were largely downregulated (summarized in Figure 7B and Figure 7—figure supplement 1A and B). These data suggest that in vivo the RPE cells are reducing oxidative phosphorylation and meeting their energy demands by increasing glycolysis. To test if RPE cells reduce oxidative metabolism in response to hypoxia, we monitored oxygen consumption rates in cultured human RPE cells (provided glucose, pyruvate and glutamine) after the cells were exposed to hypoxic conditions (3% O2) for 72 hr compared to a parallel control plate maintained in normoxia. Basal and maximal oxygen consumption rates were greatly reduced in cells after hypoxia exposure (Figure 7C and Figure 7—figure supplement 2A), suggesting mitochondrial respiration has been remodeled as a result of hypoxia. An internal ratio of the basal OCR divided by the proton leak (oligomycin-insensitive OCR) to normalize for potential plate-to-plate variation, showed a similar effect of hypoxia reducing oxidative capacity (Figure 7—figure supplement 2B). Treating the human RPE cells with low-dose DMOG, a HIF activator, also significantly reduced basal and maximal oxidative capacities in a dose-dependent manner (Figure 7D). In cells treated with higher doses of DMOG (>250 uM) oxidative capacity was completely lost (Figure 7D), even though there were no outward signs of toxicity or cell death. Collectively these data suggest that the oxidative metabolism of RPE cells is very sensitive to hypoxia. Figure 7 with 2 supplements see all Download asset Open asset Glucose consumption and metabolism is altered in the RPE of Vhl mutants. (A) Gene-profiling data revealed that glycogen degradation genes were significantly attenuated in Vhl mutants 3 days post induction (n=4). (B) Gene-profiling data for glucose metabolism genes were summarized by plotting changes along the glycolysis and TCA cycle pathways 3 days post induction (n=4) (red text=downregulated, blue text=upregulated). (C) Basal and maximal oxygen consumption rates (OCR) of hRPE after being exposed to 3% O2 for 72 hr or maintained at normoxia. Data are the mean plus s.e.m. (n=7–10). (D) Seahorse Flux Analysis OCR trace showing reduced OCR in hRPE cells treated with DMOG for 24 hr. Data points are the mean plus s.d. (n=6). (E and F) Changes in lactate (E) and glucose (F) levels of the media from hRPE cells, in transwells, after treatment with DMOG for 24 hr. Data are the mean plus s.e.m. (n=4). (G) Glucose levels are decreased in the sensory retina of Vhl-cKO mice 3 days post induction compared with littermate controls (n=6–10). (H) Relative glucose consumption is increased (roughly two-fold) in primary Vhl-cKO
